Disclosure of newly diagnosed HIV infection and condom use at first sex after diagnosis: a study of young Black men who have sex with men

Abstract

Background

The first purpose of the present study was to determine whether young Black men who have sex with men (YBMSM) disclose their newly diagnosed HIV infection to a male or female partner, and to determine whether this disclosure is related to condom use; the second was to identify correlates of disclosing newly diagnosed HIV infection to male sex partners, including a measure of partner-related barriers to condom use.

Methods

A sample of 125 HIV-infected YBMSM (age 15–29 years) provided cross-sectional data used for both study purposes. Recruitment occurred in a mid-size city in the southern US experiencing inordinately high prevalence and incidence rates of HIV among YBMSM. Significance was defined by an α level of <0.05.

Results

Eighty-eight YBMSM (70.4%) indicated disclosing their newly diagnosed HIV status to the first male partner they had sex with after being diagnosed. Of these, nine (9.1%) reported that condoms were not used during ensuing sex with that partner. However, of the men not disclosing, 27.0% reported not using condoms for ensuing sex (P = 0.009). Similar findings were observed relative to sex with females (P = 0.057). Regarding the second study purpose, in addition to a protective effect of advancing age, men scoring at or above the median on a measure of partner-related barriers to condom use were 2.4-fold more likely to not disclose compared with men scoring below the median (P = 0.04).

Conclusion

For YBMSM, a beneficial counselling objective relative to disclosing newly diagnosed HIV may be to help men resolve perceptions of partner-related barriers to condom use.

Introduction

Despite accounting for an estimated 2% of the US population, gay and bisexual males account for an estimated three-quarters of all new HIV infections.^1–3 Young Black men who have sex with men (YBMSM) have experienced a far more rapid escalation of the HIV/AIDS epidemic than who have sex with men (MSM) who identify as White or Latino or any other subpopulation in the US.^3–7 YBMSM have a one in four chance of becoming infected with HIV by the time they reach 25 years of age.⁸ This elevated prevalence of HIV among YBMSM creates sexual networks of high seroprevalence that, in turn, magnifies the odds of HIV acquisition for HIV-uninfected men in these networks. Thus, a primary public health strategy is to promote safer sex for people living with HIV, with YBMSM being a high priority.⁵

Evidence suggests that condomless anal sex (CAS) is continually practiced among MSM who have been diagnosed with HIV.⁹ Lack of disclosing HIV status to sex partners may be common, with one large-scale study suggesting that 42% do not disclose.¹⁰ Two separate studies each found that approximately 13% of the HIV-infected MSM sampled did not disclose and did have subsequent CAS.^9,10 This lack of disclosure before condomless sex prevents their partners from informed sexual decisions and perhaps adopting protective sexual behaviours if they are serodiscordant.¹⁰ Thus, this failure to disclose HIV status has been associated with CAS,¹¹ although this association was not found in a similar study.¹⁰ This phenomenon is also true for MSM newly diagnosed with HIV, because a significant number continue to have CAS.¹² Findings from past research suggest that depressive symptoms, denial and lack of disclosure self-efficacy can decrease the likelihood of MSM newly diagnosed with HIV to disclose their status to potential sexual partners.^13–15

Black MSM are less likely than White MSM to disclose their HIV status; however, when Black MSM do disclose their status, they are less likely to engage in CAS.^16,17 Thus, for HIV-infected YBMSM, a key factor in reducing transmission is disclosure. Evidence suggests that Black MSM very often have CAS with partners of an unknown HIV status, indicating that many men are not disclosing or discussing HIV status.¹⁸

The extent to which disclosing HIV status affects condom use has yet to be fully elucidated, with some studies suggesting no effect^10,19,20 and others suggesting that condom use may be less likely after disclosure.^21–25 Unfortunately, research focusing specifically on the population at greatest risk (YBMSM) relative to disclosure and condom use has not been published. Further, past research has neglected to focus on the disclosure and/or condom use immediately after testing positive for HIV. This time period is important because antiretroviral therapy (ART) may not be initiated yet and, even when it has been, viral loads may still be high enough for transmission to be possible. Accordingly, the present study investigated whether YBMSM disclose their newly diagnosed HIV infection to a male or female sex partner, and determined whether this disclosure was related to subsequent condom use.

Further, past studies of disclosing HIV status among MSM have neglected to assess whether men’s anticipated barriers to negotiating condom use with partners (possibly occurring as a consequence of the disclosure) may be an important correlate of disclosure. Thus, a second aim of the present study was to identify correlates of disclosing newly diagnosed HIV infection to male sex partners, including a measure of partner-related barriers to condom use. Past evidence has broadly implicated partner-related barriers as being a primary deterrent to condom use among young Black males, including those having sex with males.^26–29

Methods

Study sample

A convenience sample of 609 YBMSM was recruited for participation from a National Institutes of Health (NIH)-funded randomised controlled trial (RCT) of a safer sex intervention program designed specifically for this population. For the present study, only baseline data (collected before randomisation and intervention) are presented. Participant recruitment occurred in a federally supported clinic designated for the diagnosis and treatment of HIV and other sexually transmitted infections. The clinic was located in a mid-size southern city where incidence rates of HIV are particularly high. Inclusion criteria were as follows: (1) assigned male at birth; (2) self-identification as Black or African American; (3) aged 15–29 years; (4) attending the clinic to be tested for HIV or other sexually transmissible infections (STIs); (5) having engaged in anal sex with a male partner at least once in the past 6 months; and (6) the ability to speak and comprehend English.

All age-eligible Black men were approached in the clinic and asked about their interest in volunteering for an HIV prevention study. Those expressing interest were screened for eligibility. In all, 789 men were screened; of these, 623 were eligible. After being offered the opportunity to enrol, 14 declined, yielding an overall participation rate of 97.7% and a sample size of 609 YBMSM. Of these 609 YBMSM, 140 (23.0%) were HIV-infected at enrolment. All study procedures were approved by the institutional review boards of the University of Mississippi Medical Center, the Mississippi State Department of Health and the University of Kentucky.

Study procedures

After providing written informed consent (or parental consent for those under 18 years of age), participants completed an online questionnaire using Qualtrics (Provo) in a private office not physically connected to the clinic. The self-administered questionnaire had an optional audio component in the event that men experienced literacy issues.

Self-reported measures

The questionnaire items that apply directly to the research question were:

1. Think about the last time you had sex with a male partner after you tested positive for HIV. When you first had sex with him after testing positive did you discuss your HIV status with him?

2. When you first had sex with a male partner after testing positive for HIV and discussing your HIV status, did you use a condom?

Response options for the first question were ‘no’, ‘yes, but less than 30 s’. ‘yes, but less than 2 min’ and ‘yes, but less than 4 min’. Response options for the second question were ‘yes’ and ‘no’. These same two questions were also asked of HIV-infected men relative to sex with females. Fifteen of the 140 HIV-infected men opted not to answer these questions, leaving an analytical subsample of 125 men.

To address the second aim of the study, measures were selected based on past studies,^13–15 as well as the addition of a measure assessing perceived partner-related barriers to condom use. In addition to age, selected single-item correlates of disclosing were: (1) having any education beyond high school; (2) being single or never married; (3) identification as gay to other males; (4) the question, ‘In the last 12 months, how often have you been in a sexual situation where bringing up condoms would hurt the relationship at the time?’; and (5) whether men had more than one male anal sex partner in the past 90 days.

Two scale measures were also assessed. One was a previously validated measure of internalised homophobia, namely a seven-item scale developed by Meyer.³⁰ The scale obtained an adequate inter-item reliability coefficient of 0.81. The second measure was a five-item scale assessing perceptions of partner-related barriers to condom use. This measure is a subscale of an adapted version of the Condom Barriers Scale and has been used previously with a clinic-based population, including young Black males.³¹ Items were presented as statements, such as ‘If I asked my main male sex partner to use a condom, he might think I was cheating’. Men were provided with response options using a visual analogue scale ranging from 1 ‘strongly disagree’ to 5 ‘strongly agree’. The scale obtained an adequate inter-item reliability coefficient of 0.79.

Data analysis

To address the first aim of the study, Chi-squared tests were used to determine associations between disclosure and condom use. For the second aim, the two distributions stemming from the scale measure were evaluated for normality and both were highly skewed. Consequently, they were each dichotomised by a median split. Chi-squared tests were used to determine bivariate associations between the selected correlates and whether men disclosed to their male sex partners. In addition, independent-groups t-tests were used to evaluate the bivariate associations between age and this outcome. Bivariate associations within a screening level of significance (P ≤ 0.20) were then entered into a multiple logistic regression model using forward Wald entry (a stepwise method was used because this was not a hypothesis-driven investigation; instead, it was designed to identify possible correlates of non-disclosure and the possible role of partner-related barriers in subsequent condom use). Significance at the multivariable level was defined as P < 0.05.

Results

Demographic characteristics

The mean (± s.d.) age of the sample was 24.0 ± 3.1 years. An average monthly income of less than US$1000 was reported by 49.6% of study subjects. The majority of participants (49.6%) reported they were currently employed. Nearly two-thirds of the sample (64.0%) reported having education beyond high school graduation, and 32.8% reported current enrolment in a school or college.

Disclosure and condom use: male sex partners

Of the 125 men, 88 (70.4%) indicated that they had at least a 30-s discussion (regarding the test results) with the first male partner they had sex with after being diagnosed with HIV. Most of these men (n = 36) indicated that the discussion lasted more than 2 min; however, 30 stated that the discussion lasted less than 30 s, 11 indicated that it lasted between 30 s and 1 min, and 11 indicated that it lasted between 1 and 2 min. The remaining 37 men (29.6%) indicated not disclosing their status. Of the 88 men having this discussion, nine (9.1%) reported that condoms were not used when having sex with that partner. In contrast, of the 37 men who did not disclose their status, 10 (27.0%) reported not using condoms for ensuing sex with those partners. This difference was significant (P = 0.009).

Disclosure and condom use: female sex partners

Of the 15 HIV-infected men responding to two questions relevant to disclosure and condom use with female partners, nine indicated having at least a 30-s discussion with the first female partner they had sex with after being diagnosed with HIV. Most of these men (n = 4) indicated that the discussion lasted more than 2 min; however, one stated that the discussion lasted less than 30 s, two indicated that it lasted between 30 s and 1 min, and two indicated it lasted between 1 and 2 min. Of these nine men, five (55.6%) reported that condoms were not used the next time sex occurred with that partner. In contrast, six men did not discuss or disclose their test results and each of these six men (i.e. 100%) reported not using condoms for ensuing sex with those partners. This difference was not significant (P = 0.057).

Correlates of disclosure to male sex partners

The observed bivariate associations between selected correlates and disclosure to male sex partners are given in Table 1. It can be seen that five correlates obtained screening significance. Men were more likely to not disclose their newly diagnosed HIV infections to male sex partners if they: (1) had not received education beyond high school; (2) scored at or above the median on the measure of internalised homophobia; (3) scored at or above the median on the measure of perceived partner-related barriers to condom use; (4) provided any indication of having past male sex partners with whom condom use would hurt the relationship; and (5) having more than one male sex partner in 90 days preceding study enrolment.

Table 1.

Bivariate associations between selected correlates and not disclosing HIV status to sex partners after being newly diagnosed

Correlate	No. men not disclosing/total no. men in category (%)	PR	P-value
Education beyond high school		1.56	0.10
Yes	19/78 (24.4)
No	18/47 (38.3)
Relationship status is single or never married		0.77	0.37
Yes	12/48 (25.0)
No	25/77 (32.5)
Identify as gay to other males		0.90	0.71
Yes	24/78 (30.8)
No	13/97 (27.7)
Internalised homophobia		1.61	0.08
Below median	16/69 (23.2)
Above median	21/56 (37.5)
Five-item measure of partner-related barriers to condom use		1.98	0.02
Below median	12/61 (19.7)
Above median	25/64 (39.1)
Recently been in a situation when condom use would hurt relationship		1.45	0.07
No	25/94 (26.6)
Yes	12/31 (38.7)
Multiple anal sex partners in the past 90 days		1.70	0.07
No	11/51 (21.6)
Yes	25/68 (36.8)

n = number of men not disclosing, n = total number of men represented in the indicated category of the measure; PR, prevalence ratio

Table 2 lists the significant adjusted odds ratios obtained from the regression model. As shown, only two variables obtained significance. Men scoring at or above the median on the dichotomous measure of partner-related barriers were 2.4-fold more likely to not disclose compared with men scoring below the median. In addition, a protective effect occurred for age, with an observed 15% decrease in the odds of not disclosing for each advancing year of age.

Table 2.

Adjusted odds ratios (aORs) from the final logistic regression model

Correlate	aOR (95% CI)	P-value
Greater than median on partner-related barriers scale	2.40 (1.04–5.55)	0.04
Age	0.84 (0.74–0.96)	0.016

CI, confidence interval

Discussion

Findings from this sample of 125 YBMSM living with HIV in an area of the southern US suggest that as many as one of every three may not immediately disclose their newly discovered HIV diagnosis to male partners who they first have sex with after being diagnosed. Further, because those not disclosing were about threefold more likely to have CAS with these partners, the findings also convey the role of disclosing relative to condom use and thus preventing transmission. Because these men may not have yet been placed on ART or treatment may have been initiated only recently, their viral loads may have posed a dangerously high risk of male-to-male transmission, especially if the partner was receptive.

From an intervention perspective, the findings provide guidance to disease intervention specialists (DIS; people conducting contact tracing), HIV counsellors and other healthcare providers working in settings that diagnose and treat HIV. Specifically, men’s perceptions that any partner-related barriers to condom use may exist could be a substantial detriment to disclosure, thereby suggesting that counselling designed to overcome these perceptions may promote subsequent disclosure. It is important to consider that the partner-related barriers are perceptions and, as such, may be predicated on faulty assumptions about partners. Clearly, a large number of considerations most likely affect the ultimate decision to disclose. Thus, placing the study findings in perspective, the apparent strength of this scale measure to predict disclosure suggests that some newly HIV-diagnosed men may be reticent to suggest condom use as one of many considerations they entertain when deciding to disclose their HIV diagnosis to sex partners.

A related point involves the criminalisation of not disclosing an HIV-positive status to sex partners (or injecting drug use partners) in 33 of the 50 US states.³² Whether these laws facilitate or impede disclosure is not known; however, a case can be made that the laws do act as an impediment for men who do not disclose the first time they have sex with any given partner. By disclosing after the first sex, the perception men have could be one involving criminal charges related to that first sex act. In populations being highly targeted by law enforcement efforts (in the US this occurs for young Black males), these fears may be well founded. Thus, the all-so-important counselling point for newly diagnosed YBMSM is that immediate disclosure may protect them from criminal charges.

The study findings also have implications for the HIV-uninfected men who may become sex partners of newly diagnosed YBMSM not disclosing their HIV status. For example, it may important for HIV-uninfected men to clearly ask, ‘when were you last tested for HIV and what was the result?’ Although honest answers may not always be forthcoming, the question may elicit disclosure from men who only needed an opportunity to do so.

Although statistical power was lacking, another important study finding involves the men who also had female sex partners. The P-value of 0.057 is an artefact of this low power and thus it seems quite reasonable to suggest that the same disclosure and condom use dynamics that apply to male sex partners also apply to females. Thus, once again, implications for DIS, HIV counsellors and other healthcare providers apply in that counselling sessions for YBMSM newly diagnosed should be structured in a bimodal fashion, with the second mode being devoted to resolving issues relative to disclosing newly determined HIV status to female sex partners. Of note, the small number of men having sex with females precluded more in-depth analyses (i.e. determining correlates of disclosure as was done for male partners).

Study findings pertaining to the association between disclosure and condom use are quite different than those found in a study reported by Marks and Crepaz.¹⁰ Compared with a MSM in that study,¹⁰ we observed in the present study that only approximately 30% had not disclosed their HIV status; however, our measure was quite different (disclosure to the first male partner they had sex with after diagnosis) and thus the lower value is expected. Further, compared with studies suggesting that approximately 13% of HIV-infected men do not disclose and do engage in CAS,^9,10 our finding suggests this value may be twice as high (i.e. 27%) and, unlike a study that found no association between disclosure and CAS,¹¹ the association obtained in the present study was very strong. Thus, future investigations of YBMSM and other high-risk populations of MSM should include measures designed to provide more in-depth investigation into why this association may be found for one population but not another of similar transmission risk.

Of interest, the findings suggest that the observed associations between disclosure and subsequent condom use with that partner apply with the ‘disclosure discussion’ being as brief as 30 s. How men frame this discussion in such a short time period is clearly worthy of qualitative follow-up investigation. That the disclosure can be as brief as 30 s is consistent with the prevailing practice of hooking up for sex. This sexual milieu of hooking up creates a need for disclosures of HIV-positive status to happen within a non-relational context. Therefore, counselling objectives may also attempt to instil ‘hook-up’ disclosure skills and practices in YBMSM clients.

Limitations

In addition to the use of a convenience sample and a cross-sectional study design, three other limitations apply to the findings. First, this is a secondary analysis of data for use in an RCT, thus the measures may have been less than ideal for this study of disclosure and condom use. For example, the measures of disclosure and condom use did not specify anal sex, thereby leaving open the possibility that men may have responded in thinking about oral sex partners. Second, the dichotomised measure of partner-related barriers was highly skewed, with approximately 49% of the men indicating a complete absence of any of the five barriers. Thus, men scoring at or above the median were simply reporting a degree of presence for one or more of the five barriers. Subsequent investigations should use more refined measures that create larger dispersion. In addition, we did not assess whether men abstained from sex after being diagnosed with HIV or how long this period lasted if it occurred. Finally, the very low number of men having sex with females created power issues that could reasonably cast doubt on the suggestion that a generalisable association exists between disclosure and subsequent condom use for YBMSM newly diagnosed with HIV.

Conclusion

Taken as a whole, the study findings highlight a clinical imperative to escalate counselling efforts directed towards YBMSM newly diagnosed with HIV. That disclosure of newly diagnosed HIV infection diagnosis is linked to subsequent condom use with these partners is both encouraging (regarding those who disclose) and problematic (regarding those who do not disclose). Resolving the problematic side of the association will require attention to the possibility that some YBMSM may have female as well as male sex partners. Pragmatically, the findings provide a potentially beneficial counselling objective relative to disclosing among newly diagnosed YBMSM: help men resolve perceptions of partner-related barriers to condom use.